CHANGE IN T2-FAT SATURATION MRI CORRELATES WITH OUTCOME IN CERVICAL CANCER PATIENTS

Purpose: To compare pretreatment and midtreatment tumor intensity as measured by T2 fat-saturation (T2-FS) MRI and its association with treatment response in cervical cancer patients. Methods and Materials: Weekly MRI scans were performed for brachytherapy planning on 23 consecutive patients with clinical Stage IB1 to IIIB cervical cancer treated with definitive chemoradiotherapy. These scans were performed on a 1.5-T clinical scanner using a specialized pelvic coil. Mean signal intensity from T2-FS imaging was calculated for each tumor voxel. Average tumor intensity and tumor volume were recorded pre- and midtreatment (at Weeks 0 and 4). All patients subsequently underwent routine follow-up, including periodic clinical examinations and fluorodeoxyglucose-positron emission tomography imaging. Results: Mean follow-up for surviving patients was 14.5 months. Mean tumor volume at presentation was 49.6 cc, and mean midtreatment tumor volume was 16.0 cc. There was no correlation between initial tumor volume and pretreatment signal intensity (r = 0.44), nor was there a correlation between pre- or midtreatment tumor volume with disease-free survival (p = 0.18, p = 0.08 respectively.) However, having at least a 30% drop in signal intensity from pretreatment to midtreatment was correlated with having disease resolution on posttreatment fluorodeoxyglucose-positron emission tomography imaging (p = 0.05) and with disease-free survival (p = 0.03.) Estimated disease-free survival at 22 months was 100% for patients with at least a 30% drop in tumor signal intensity compared with 33% for patients above this selected threshold (p = 0.004). Conclusions: Longitudinal changes in T2-FS tumor intensity during chemoradiation correlated with disease-free survival in cervical cancer patients. Persistently high midtreatment tumor intensities correlated with a high risk of treatment failure, whereas large decreases in tumor intensity correlated with a favorable outcome. (C) 2011 Elsevier Inc.